This research strives to develop protein-based therapeutics useful for the treatment of cancer built on the principle of eliminating key components of the solid tumor support network. In particular, this work will focus on the selective targeting of cancer-associated fibroblasts (CAFs). Cancer-associated fibroblasts have been implicated in a number of important roles in the initiation and progression of solid tumors, but selective targeting of these stromal cells with therapeutic effect has proven difficult. The most widely investigated approaches to selectively target CAFs take advantage of the nearly exclusive expression of the cell surface protease fibroblast activation protein-alpha (FAP) on the surfaces of CAFs located at the site of solid tumors. These efforts have focused on the development of antibodies and small molecules that target cells expressing FAP. However, these approaches have not yielded effective therapeutic cancer treatments. One potential reason for problems in the development of therapeutics that target CAFs via FAP is that the underlying biological role of FAP in tumor initiation and progression is still poorly understood. The work described here takes a combined approach to studying the biology of FAP and developing new cancer therapies. Three specific aims will be pursued in this project: 1) Develop binding proteins to multiple epitopes of the activated fibroblast cell surface protease fibroblast activation protein-alpha and use these binding proteins to construct multivalent antibody-based reagents. 2) Use multivalent antibody-based reagents developed in Aim 1, protease inhibitors targeting fibroblast activation protein-alpha, or combinations of the two, in order to study the effect these reagents may have on fibroblast activation protein-alpha-mediated functions in co-cultures of fibroblasts and cancer cells. 3) Use mouse models of cancer to examine the effects of reagents identified in Aim 2 on tumor growth and metastasis. The binding proteins to be generated as a part of Aim 1 will be developed using the protein engineering techniques afforded by yeast surface display. Multivalent binding proteins will combine identified binders within an antibody format. In vitro studies will primarily focus on the collective invasive behaviors of mixtures of tumor cells and CAFs and the behaviors of isolated pairs of tumor cells and CAFs using invasion assays and microwell array technology, respectively. Using binders developed in Aim 1, these studies will help elucidate which portions of FAP are responsible for invasive phenotypes and examine whether the selective downregulation of FAP expression on CAFs will inhibit these behaviors. Murine models of cancer will be used to study the therapeutic benefits of targeting FAP with multivalent binding proteins using experiments to test whether the proteins prevent tumor growth or prevent the development of immune system tolerance to a tumor. These studies will give the scientific community new tools for studying and interfering with the biological functions of FAP and new therapeutic leads for the treatment of cancer.